Background: Recent advances in computer vision, particularly in deep learning, have significantly improved object recognition capabilities in images. Among these, real-time object detection frameworks such as You Only Look Once (YOLO) have shown promise across various domains. This study explores the application of YOLO-based object detection for food identification and portion estimation, with a focus on alignment with the Swedish plate model recommended by the National Food Agency. Objective: The primary aim is to evaluate and compare the performance of three YOLO variants - YOLOv7, YOLOv8, and YOLOv9 - in detecting individual food components and estimating their relative proportions within images, based on public health dietary guidelines. Methods: A custom dataset comprising 3,707 annotated food images spanning 42 food classes was developed for this study. A series of preprocessing and data augmentation techniques were applied to improve dataset quality and model generalization. The models were assessed using standard evaluation metrics including precision, recall, mean Average Precision (mAP), and F1 score. Results: Among the evaluated models, YOLOv8 outperformed YOLOv7 and YOLOv9 in both peak precision and F1 scores. It achieved a peak precision of 82.4%, compared to 73.34% for YOLOv7 and 80.11% for YOLOv9, indicating superior accuracy in both food classification and portion estimation tasks. YOLOv8 also demonstrated stronger confidence in its predictions. However, all models exhibited some difficulty in differentiating between visually similar food items, highlighting the complexity of fine-grained food recognition. Conclusions: While YOLO-based models, particularly YOLOv8, show strong potential for food and portion recognition aligned with dietary models, further refinement is necessary. Enhancements in model architecture and training data diversity are required before such systems can be reliably deployed in health and dietary monitoring applications.